This work summarizes selected applications of artificial neural networks and related solutions such as neuro‐fuzzy within different control loops such as the network predictive control regarding energy saving. It also shows such applications in the chemical industry and points out why further research on these applications regarding the control of distillation columns might be economically promising, as the world market in the chemical industry is expected to grow from 3.47 trillion euro in 2017 to about 6.6 trillion euro in 2030, the energy consumption might also rise. As not only companies in the chemical industry set up energy saving programs to reduce energy consumption and greenhouse gases, the consumption of energy decreased since 1990 in Europe. Compared to other countries like China, the energy prices in Europe – especially in Germany – have the highest prices due to taxes and governmental issues. Thus, research about applications with artificial intelligence in energy saving holds the promise of economic gain. © 2020 Society of Chemical Industry and John Wiley & Sons, Ltd.

Artificial intelligence in process control applications and energy saving: a review and outlook

A robust double-parallel extreme learning machine based on an improved M-estimation algorithm

The traditional risk management and control mode (RMCM) in regional sites has the defects of low efficiency, high cost, and lack of systematism. Trying to resolve these defects and explore the application possibility of machine learning, a characteristic dataset for RMCM in regional sites was established. Three decision tree (DT) algorithms (CHAID, EXHAUSTIVE CHAID, and CART) and two artificial neural network (ANN) algorithms [back propagation (BP) and radial basis function (RBF)] were implemented to predict RMCM in regional sites. The results showed that in the aspects of accuracy (ACC), precision (PRE), recall ratio (REC), and F1 value, CART–DT was superior to CHAID–DT and EXHAUSTIVE CHAID–DT (E-CHAID–DT); and BP–ANN was superior to RBF–ANN. However, CART–DT was inferior to BP–ANN in ACC, PRE, REC, and F1 value. BP–ANN model is good at non-linear mapping, and it has a flexible network structure and a low risk of over-fitting. The case study of a typical county demonstration area confirmed the extensibility of the method, and the method has great potential in RMCM prediction in regional sites in the future.

Prediction Performance Comparison of Risk Management and Control Mode in Regional Sites Based on Decision Tree and Neural Network

Due to the characteristics of multi-dimensional, strong coupling and noise of complex industrial data, this paper establishes an energy efficiency recognition and diagnosis model using the multivariate nonlinear regression method to analyze the objective mathematical relationship between raw material consumption and the corresponding output in complex industrial processes. The curve estimation between raw materials and production of complex industrial processes are carried out. Then the raw materials are replaced by the calculation results of the equation with higher fitting degree. Finally, the linear regression and energy efficiency diagnosis of the substituted raw materials are carried out. The proposed model is tested and analyzed using actual industrial data of ethylene production plants. The results show the validity and applicability of the proposed model. Meanwhile, the objective mathematical relationship between the ethylene production capacity and the energy consumption of various raw materials can be recognized to adjust the proportion of raw materials in ethylene production plants, which can improve the energy efficiency and reduce raw materials waste.

Energy Efficiency Recognition and Diagnosis of Complex Industrial Processes using Multivariate Nonlinear Regression Method

For the conventional radial basis function (RBF) neural network, there are many problems like uncertain nodes in the hidden layer, sensitivity to initial centers and slow convergence speed, etc. This paper proposes an RBF neural network model based on the fuzzy C-means method (FCM-RBF), with each cluster center obtained by fuzzy C-means clustering. And weights between the hidden layer and the output layer are trained by the gradient descent method based on error back-propagation (BP). The proposed method overcomes the sensitivity of the data center for traditional RBF model, determines optimally the number of nodes in the hidden layer of RBF neural network, and improves the network training speed and precision. Finally, the proposed method is applied in the production capacity forecast of the ethylene plants. The production statuses of ethylene plants of different technologies or different scales are analyzed and predicted to guide the ethylene production and improve energy efficiency. The empirical results demonstrate the effectiveness and practicability of the proposed

Ethylene plants production capacity forecast based on fuzzy RBF neural network

Extreme learning machine (ELM) with a hidden layer has been successfully applied in many fields In order to enhance the generalization performance of ELM, an enhanced ELM with a double parallel structure (DP-ELM) was proposed In the proposed DP-ELM, a special connection between the input nodes and the output nodes was built With the connection between the hidden nodes and output nodes, the output nodes of DP-ELM can receive the information from hidden nodes With the special connection, the self-information from inputs can be directly passed to the output nodes To verify the performance of the proposed DP-ELM, simulations are carried out using 5 regression datasets, including two process modeling applications Simulation results show that, compared with ELM, the proposed DP-ELM could achieve higher accuracy better stable ability

An enhanced extreme learning machine with a double parallel structure and its application to modeling complex chemical processes

Comprehensive energy modeling and optimization play a key role in sustainable development of complex petrochemical industries. However, it is difficult to make effective energy modeling and optimization due to the characteristics of uncertainty, high nonlinearity, and with noise of modeling data from the practical production. To deal with this problem, a novel energy modeling and efficiency optimization method using a novel extreme learning fuzzy logic network (ELFLN) is proposed. In the proposed method, Mamdani type fuzzy inference system (FIS) and multi-layer feedforward artificial neural network (MLFANN) are adopted. First, the fuzzy inference replaces the hidden layers of artificial neural network (ANN). Then the proposed framework takes fuzzy membership degrees instead of precise values as the output. Meanwhile, an extreme learning algorithm based on Moore-Penrose Inverse is utilized to train the network efficiently. Three levels of energy efficiency of “low efficiency, median efficiency and high efficiency” can be effectively achieved using the proposed method. For inefficiency samples, valid slack variables are predicted for finding the direction of improving the efficiency. The energy efficiency optimization performance and the practicality of the proposed method is confirmed through an application of China ethylene industry. Finally, the energy saving potential is indicted as 8.82% and practical ethylene production can be guided by the result of the demonstration analysis.

Energy modeling and efficiency optimization using a novel extreme learning fuzzy logic network

Building a robust and accurate energy analysis model is considered as an important issue in the field of petrochemical industries. Under the circumstance of small samples, the accuracy of the energy analysis model is unacceptable. In order to solve this problem, a novel noise injection integrated with extreme learning machine based nonlinear virtual sample generation method is proposed. Through injecting noise in the output matrix of the hidden layer of ELM, a virtual information matrix that is different the original one generated using the original small dataset can be obtained. Then the newly generated information matrix is adopted to produce good-quality virtual samples for supplement knowledge for small samples. To authenticate the effectiveness of the proposed method, a standard trigonometric function is first selected; and then the proposed method is developed as an energy analysis model for an ethylene production process. Simulation results indicate that good virtual samples can be generated using the proposed method, and the accuracy of the energy analysis model is much improved with the aid of the newly generated virtual samples. The proposed method will effectively help production departments of petrochemical industries set more suitable targets of energy consumption and make better use of available resources.

A novel nonlinear virtual sample generation approach integrating extreme learning machine with noise injection for enhancing energy modeling and analysis on small data: Application to petrochemical industries

The power system measurement data has high-dimensional features and strong noise, which is difficult to be directly used for intrusion detection Traditional feature extraction and selection methods take the feature processing as a preprocessing step and perform separately from the model training, which makes the features not well adapted to the model Therefore, we propose a novel wrapped feature selection framework based on the Las Vegas algorithm, which can improve the detection accuracy by strengthening the coupling between the feature selection and the model training The Las Vegas algorithm can evaluate a feature subset through a specified model In this paper, a heuristic thinking is integrated into the Las Vegas algorithm, which greatly improves the search performance of the original method Finally, the proposed framework is examined on the IEEE 14-bus, 39-bus and 57-bus test systems for two attacks and the experimental results proves the effectiveness and stability of the proposed framework

Geng Zhiqiang

Papers

Ethylene plants production capacity forecast based on fuzzy RBF neural network

An enhanced extreme learning machine with a double parallel structure and its application to modeling complex chemical processes

Energy modeling and efficiency optimization using a novel extreme learning fuzzy logic network

A novel nonlinear virtual sample generation approach integrating extreme learning machine with noise injection for enhancing energy modeling and analysis on small data: Application to petrochemical industries

Power Systems Intrusion Detection Using Novel Wrapped Feature Selection Framework